in vitrorelease of model compounds with different ... · the procedure for preparation of...

Acta Poloniae Pharmaceutica ñ Drug Research, Vol. 70 No. 6 pp. 1051ñ1063, 2013 ISSN 0001-6837Polish Pharmaceutical Society

Microspheres are a useful type of delivery sys-tem for administration of drugs, since at one fellswoop they can be used to encapsulate, protect andcontrol release of a wide variety of drugs (1-4). Theyare especially useful due to that they are easilyadministered by injection. Biodegradable micro-spheres do not require surgical removal after drugexhaustion (5, 6). The drug release rate from bio-degradable microspheres can be controlled bymanipulation with the particle size (7-9) the polymerdegradation rate, the polymer degradation and ero-sion mechanism as well as other factors (10, 11).Tamada and Langer (12) classified degradable poly-mers into surface eroding and bulk eroding ones.The bulk erosion would be expected for polyesterswhereas polyanhydrides exhibit a surface erosionprofile controlled by the hydrophobic nature of thepolymer backbone. Surface eroding polyanhydridesmay simplify the drug release kinetics because waterpenetration into the microsphere interior is mini-mized, and the drug release rate becomes dependentpredominantly on the polymer erosion rate (8, 13,

14). Poly(ester-anhydride)s combine the individualproperties of two different classes of biodegradablepolymers. The differences in degradation behavior(depending on the poly(ester-anhydride) copoly-mers ratio) can be used to influence release profilesof encapsulated bioactive compound (11, 15).However, drug release and polymer degradation aresometimes not correlated, depending on the relativehydrophobicity and compatibitity of the co-monomers as well as the drug loaded. Thus, theinteraction between the polymer and the drug ofteninfluenced the mechanism of drug release due tonon-uniform drug distribution (8). One of criticalfactors determining drug release rate is microspheresize. In many cases, smaller microspheres releasethe active compound faster due to increased surfacearea/volume ratio (7). However, the effects ofmicrosphere size can be more complex. The smallmicrospheres used to solidify more quickly andexhibit more homogeneous drug distribution thanlarge ones (especially when drug and polymermatrix differ in hydrophilicity) resulting in slower

PHARMACEUTICAL TECHNOLOGY

IN VITRO RELEASE OF MODEL COMPOUNDS WITH DIFFERENTHYDROPHILICITY FROM POLY(ESTER-ANHYDRIDE) MICROSPHERES

KATARZYNA JASZCZ*

Department of Physical Chemistry and Technology of Polymers, Silesian University of Technology, 44-100 Gliwice, M. Strzody 9, Poland

Abstract: Poly(ester-anhydride) microspheres were prepared using emulsion solvent evaporation techniquefrom two copolymers obtained by polycondensation of sebacic acid (SBA) and oligo(3-allyloxy-1,2-propylenesuccinate) terminated with carboxyl groups (OSAGE). The SBA content in copolymers was 90 or 70 w/w %,respectively. The size of microspheres obtained was in the range of 2-4 µm (small microspheres) or 12-31 µm(large ones) and depended on stirring conditions used in emulsion formulation process, as well as on concen-tration of polymer solution used. Poly(ester-anhydride) microspheres were loaded with three model compounds(rhodamine B, p-nitroaniline and piroxicam) with different water solubility. The loading efficiency was depend-ent on kind of model compound, polymer composition and size of microparticles. Each of three model com-pounds also exhibited different release profiles. Rhodamine was released very quickly. The release of p-nitroaniline was more prolonged, but showed only little dependence on microsphere size and polymer compo-sition, whereas the piroxicam-loaded microspheres exhibited the most interesting release profile, showing thatrelease rate as well as transport mechanism can be adjusted by changing microsphere size and poly(ester-anhy-dride) compositions.

Keywords: biodegradable microspheres, emulsion solvent evaporation, poly(ester-anhydride), release kinetic,drug delivery

1051

* Corresponding author: e-mail: [email protected]

1052 KATARZYNA JASZCZ

releasing rate. Specific physicochemical characteris-tics of a drug, such as molecular weight, molecularvolume and polarity, are also expected to affect theinteraction between drugs and polymers, both dur-ing formulation and release experiments, and mayalso influence the release mechanism (8, 16-18).

Recently, we have obtained functionalpoly(ester-anhydride) microspheres with relativelynarrow size distribution by emulsion solvent evapo-ration technique (19). In this paper, we present thestudies on loading the microspheres with modelcompounds with different hydrophilicity and theirrelease characteristics.

The aim of this work was to evaluate the influ-ence of polymer composition and size of micropar-ticles on loading efficiency and release profiles ofmodel compound with different hydrophilicity.Three low molecular weight model compounds: rho-damine B (water solubility 8 mg/mL), p-nitroaniline(water solubility 1 mg/mL) and piroxicam (watersolubility 0.1 mg/mL) were chosen to investigate therole of different ìcompatibilityî of loaded com-pound with polymer matrix on the drug loading andrelease mechanism.

EXPERIMENTAL

Materials

Succinic acid 99% (Aldrich), sebacic acid 98%(ACROS Organic), allyl glycidyl ether 99%(Aldrich) and acetic anhydride (POCH S.A.),poly(vinyl alcohol) (MW = 88000 g/mol, 88%hydrolyzed, ACROS Organic), rhodamine B and p-nitroaniline (Fluka), piroxicam (Aldrich) and phos-phate buffer solution pH = 7.41 (LabStand PoznaÒ)were used as supplied. Solvents were purifiedaccording to known procedures.

Synthesis of poly(ester-anhydride)sPoly(ester-anhydride)s were synthesized in

two-step reaction from sebacic acid (SBA) andoligo(3-allyloxy-1,2-propylene succinate) terminat-ed with carboxyl groups (OSAGE), according toprocedure described earlier (20). Copolymers pur-posed for the preparation of microspheres in thiswork were synthesized with use of 70 or 90% w/wof SBA. SBA and OSAGE (total amount 10 g) wererefluxed in 100 mL of acetic anhydride, under nitro-gen for 30 min, to yield mixed prepolymers. Anexcess of acetic anhydride and acetic acid formed asa by-product were next removed under vacuum. Inthe second step, mixed prepolymers were condensedto yield poly(ester-anhydride)s. Polycondensationwas carried out at 150OC for 2 h under high vacuum

(finally 0.01 mm Hg). The poly(ester-anhydride)sobtained were dissolved in methylene chloride, pre-cipitated in petroleum ether and dried at 40OC undervacuum. The polymers obtained were stored at -18OC prior to be used for microspheres formulation.

Microsphere preparation

The poly(ester-anhydride)s obtained were for-mulated into unloaded (blank) and model compoundloaded microspheres using emulsion solvent evapo-ration technique. Parameters of the process such as:concentration of polymer solution and speed ofhomogenizer were chosen basing on the researchdescribed previously (19) to obtain small or largemicroparticles.

Blank microspheresThe polymer solution in methylene chloride

(20 ml, concentration 10 or 50 mg/mL) was emulsi-fied in 400 mL of aqueous solution (0.5 % w/w) ofpoly(vinyl alcohol) (PVA), using ULTRA-TUR-RAX T18 homogenizer (3000 or 18000 rpm), for 30s. The emulsion was then stirred with magnetic stir-rer at 1100 rpm at room temperature for 3 h to evap-orate the organic solvent. After that, microsphereswere collected by centrifugation at 5000 rpm for 5min, washed 3 times with distilled water, lyophyl-ized and stored in freezer. The yield of microparti-cles obtained were ca. 40% for small particles andabove 70% for large ones.

Model compound loaded microspheres The procedure for preparation of p-nitroaniline

or piroxicam loaded microspheres was similar tothat used for preparation of blank particles. p-Nitroaniline or piroxicam (10% w/w in respect tothe mass of polymer) was dissolved in methylenechloride polymer solution and then the organicphase was emulsified in aqueous PVA solution. Thesolidification and isolation of drug loaded micros-pheres were performed similarly to blank ones.

For encapsulation of water soluble rhodamineB, water-oil-water (W/O/W) technique was used.Rhodamine B (3% w/w with respect to the mass ofpolymer) was dissolved in 0.8 mL of distilled water.The aqueous solution was then emulsified in 20 mLof polymer solution in methylene chloride, to formW/O inner emulsion. The inner emulsion was nextemulsified in 400 mL of aqueous solution of PVA,forming W/O/W emulsion. The inner emulsion wasobtained using ULTRA-TURRAX T18 homogeniz-er at 9000 rpm, for 30 s. The outer emulsion wasobtained using 3000 rpm when large microsphereswere prepared or 18000 rpm when small microsphe-

In vitro release of model compounds with different hydroplilicity... 1053

res were prepared. The solidification and isolationof microspheres were performed similarly to blankones.

Characterization of the particles

Shape, size distribution of microspheres wereestimated using optical microscope (OM) DELTAOptical ME 100 and PHAMIAS 2003 v.1.3 B soft-ware, as described earlier (19). The morphologicalcharacterization of microspheres was carried out ona TESLA BS 340 scanning electron microscope(SEM). The differences in thermal properties ofpoly(ester-anhydride) blank or loaded microsphereswere observed by differential scanning calorimetry(DSC), using 822e DSC Mettler Toledo apparatus.Samples were tested at the temperature ranging from-70 to 250OC at a heating rate of 10OC/min.

Hydrolytic degradation

Hydrolytic degradation of blank microsphereswas performed in phosphate buffer solution of pH =7.41 (PBS) at 37OC. Microspheres (ca. 0.2 g) wereplaced in vials containing 20 mL of PBS. The vialswere incubated at 37OC for various time (1 h to 14days). After incubation, the samples were separatedby centrifugation, washed with water, dried in vacu-um at 50OC and weighted to estimate weight loss.

In vitro model drug release

Drug release studies were performed duringtheir hydrolytic degradation in PBS at 37OC understatic conditions. Microspheres (~5 mg) containingmodel compound were placed in vials and suspend-ed in 1.5 mL of PBS. The vials were incubated at37OC. At certain time point (1 h to 30 days), thesamples were centrifuged, 1 mL of supernatant wasremoved and the medium was supplemented withfresh PBS. Vials were briefly vortexed to resuspensethe microspheres. Concentration of model com-pounds in the supernatant was determined by meas-uring the UV absorbance at the λmax = 553 nm forrhodamine B, 377 nm for p-nitroaniline and 394 nmfor piroxicam. All these experiments were per-formed in triplicate.

Cumulative release (Su) of model compoundwas calculated in respect to mass of microspheres,according to eq. (1) and (2):

Su = ΣmSMn-1 + 1.5mSMn (1)mSMn = (C ∑ r)/mm (2)

where: Su ñ cumulative amount of model compoundreleased from 1 mg microspheres within definedtime [µg/mg]; mSMn ñ the mass of model compoundin the n-th sample of supernatant (taken after a spec-ified time) in respect to 1 mg of microspheres

[µg/mg]; C ñ concentration of the model compoundin a buffer solution in the n-th sample [µg/mL]; r ñdilution of the buffer solution used for the analysisof UV; mm ñ weight of microspheres [mg]

Absorbance spectra of model compounds,obtained from fresh solution and supernatant sam-pled at various times during the release experiment,were identical in shape indicating that there was nodegradation of model compounds during the releaseperiod.

Furthermore, the model compound releasemechanism was determined by fitting the experi-mental data to Korsmeyer-Peppas equation model(21, 22). This model is used to analyze release ofpharmaceutical polymeric dosage forms when therelease mechanism is not well known or when morethan one type of release phenomena are involved(18).

Drug loading efficiency

Total amount of model compound contained inmicrospheres was obtained in separate experimentby accelerated degradation in PBS at 70OC. Thesame conditions of the hydrolytic degradation werealso applied for unloaded microspheres, to obtainblank test in order to verify whether the UV spec-trum of the post-degradation solution of blankmicrospheres do not overlap with the UV spectra ofmodel compounds. Post-degradation solutions ofunloaded microspheres do not show the absorbancewithin the range of bands characteristic for modelcompounds.

The actual loading of model compound (LA)and encapsulation efficiency (EE) were calculatedfrom the weight of the initial ñ drug loaded micros-pheres and the amount of model compounds usedand incorporated, according to equation (3) and (4).

LA = (mSM / mm) (3)% EE = (LA / LTh) ◊ 100 (4)

where: LA ñ actual loading of model compound[µg/mg]; mSM ñ the weight of model compound [µg]encapsulated in microspheres; mm ñ the weight ofmodel compound loaded microspheres [mg]; LTh ñtheoretical loading of model compound for piroxi-cam and p-nitroaniline, LTh = 100 µg/mg, for rho-damine B LTh = 30 µg/mg.

LA of piroxicam and p-nitroaniline were alsodirectly determined by dissolving 1 mg of loadedmicroparticles in chloroform and subsequent deter-mination of amount of piroxicam and p-nitroanilinein organic solutions by UV.

LA values determined in PBS solutions afteraccelerated degradation of microspheres (Table 1)and these determined in organic solutions were sim-


ilar to Su values obtained after relatively longrelease period.

RESULTS AND DISCUSSION

Functional poly(ester-anhydride)s (PSAGESB)with allyl pendant groups (Fig. 1) were obtained bypolycondensation of sebacic acid (SBA) andoligo(3-allyloxy-1,2-propylene succinate) (OSAGE).The detailed characteristics of poly(ester-anhy-dride)s were described earlier (20, 23). Copolymerspurposed for the preparation of microspheres in thiswork were synthesized with use of 70 or 90% w/wof SBA in the reaction mixture.

Formulation of poly(ester-anhydride) micro-

spheres

Microspheres were prepared using emulsionsolvent evaporation technique (O/W). Poly(vinylalcohol) (PVA) was used as emulsion stabilizingagent. Incorporation of rhodamine B (the mosthydrophilic substance) into microspheres requiredthe use of inner W/O emulsion prepared by disper-sion of an aqueous rhodamine B solution in polymersolution in methylene chloride. The inner emulsionwas then dispersed in 0.5% of aqueous PVA solu-tion to form W/O/W emulsion. Such method(W/O/W) is commonly used for encapsulation ofhydrophilic substance into microspheres (1). During

Table 1. Microsphere size, actual and theoretical loading and encapsulation efficiency of various microsphere samples and various modelcompounds.

Kind of Dn ± S LA ± SDa LTh EE ± SDa

model compound microspheres (Dv/Dn) [µg/mg] [µg/mg] [%]

PSAGESB90 17.75 ± 5.04 - -large (1.37)

PSAGESB70 24.87 ± 7.17large (1.22)

- -Blank particles

PSAGESB90 2.57 ± 0.640

small (1.19) - -

PSAGESB70 1.98 ± 0.45small (1.16)

- -

PSAGESB90 27 ± 5.98large (1.48)

9.4 ± 0.8 31.3 ± 2.7

PSAGESB70 31.44 ± 7.51 large (1.52)

4.1 ± 0.7 13.7 ± 2.3Rhodamine B

PSAGESB90 3.04 ± 0.9730

small (1.29) 10.9 ± 1.2 36.3 ± 4.0

PSAGESB70 2.39 ± 0.11small (1.21)

6.1 ± 0.6 20.3 ± 2.0

PSAGESB90 15.59 ± 4.95 large (1.32)

61.9 ± 3.3 61.9 ± 3.3

PSAGESB70 23.77 ± 8.89large (1.33)

56.3 ± 2.3 56.3 ± 2.3p-Nitroaniline

PSAGESB90 2.05 ± 0.54100

small (1.22) 12.1± 1.2 12.1 ± 1.2

PSAGESB70 2.32 ± 0.65small (1.27)

48.8 ± 3.1 48.8 ± 3.1

PSAGESB90 12.64 ± 4.77 large (1.44)

16.2 ± 0.9 16.2 ± 0.9

PSAGESB70 14.19 ± 5.29large (1.37)

29.3 ± 1.0 29.3 ± 1.0Piroxicam

PSAGESB90 4.01 ± 1.16100

small (1.25) 7.3 ± 0.7 7.3 ± 0.7

PSAGESB70 2.47 ± 0.78small (1.27)

29.7 ± 1.0 29.7 ± 1.0

EE ñ encapsulation efficiency ( % ± SDa); LA - actual model compound loading (µg/mg ± SDa); LTh - theoretical model compound loading;aStandard deviation, calculated on three different batches


the preparation of microspheres such parameters asconcentration of polymer solution and speed ofhomogenizer were changed to prepare small parti-cles with diameter within the range Dn = 1.9ñ4.2 µmor large particles with Dn = 12.4ñ31.4 µm, respec-tively. From each poly(ester-anhydride)s small andlarge microspheres (Fig. 1) loaded with one of threemodel compounds with different hydrophilicity aswell as unloaded (blank) microspheres were pre-pared (Table 1).

Microspheres with the highest diameters wereobtained from PSAGESB70 solution with concen-tration of 50 mg/mL, whereas the smallest particleswere prepared from PSAGESB70 solution with con-centration of 10 mg/mL. The size of microparticleswas dependent also on the kind of model compoundloaded, besides of basic factors of preparation con-

ditions. The sizes of microspheres containing p-nitroaniline were most similar to blank microparti-cles, probably due to its relatively good compatibil-ity with the polymer matrix. The presence of p-nitroaniline did not influence solidification ofmicrospheres during the solvent evaporation fromthe dispersion. The compatibility of p-nitroanilinewith poly(ester-anhydride)s was confirmed by ther-mal analysis. DSC thermograms were performed forblank and model compounds loaded, large micros-pheres. The presence of model compounds inmicrospheres caused decreasing of melting temper-ature (Tm) of poly(ester-anhydride)s as well asreducing of the heat of fusion (∆Hm), compared toTm and ∆Hm registered for blank microparticles.These changes were least significant for micros-pheres loaded with p-nitroaniline (Table 2).

Figure 1. The scheme of preparation of model compounds loaded poly(ester-anhydride) microspheres

Figure 2. Scanning electron microphotographs of (A) large PSAGESB70 microspheres loaded with piroxicam (B) small PSAGESB90microspheres loaded with rhodamine B


Piroxicam and rhodamine B did not affect thedimension of small microspheres, but influenced thesize of large ones. Large microspheres containingpiroxicam were slightly smaller than respectiveblank ones, whereas large microspheres loaded withrhodamine B were, in turn, slightly larger than theblank ones (Table 1). The kind of the model com-pound had also an impact on the surface morpholo-gy of microspheres. The surface of microspherescontaining the active substances (observed by SEM)were not smooth and uniform (homogeneous) incontrast to surface of blank particles. Active sub-stances were visible on the surface of microspheresin form of cambers coated with polymer or uncoat-ed crystals. Piroxicam crystals (as the needles com-pletely uncovered by polymer) were very well visi-ble on the surface of microspheres (Fig. 2A). Lesspiroxicam crystals were observed at the surface ofsmall microspheres than at the surface of large ones.Rhodamine B (in form of many pins or camberscovered by polymer) was visible at the surface ofsmall and large microspheres with similar surfacedensity (Fig. 2B). p-Nitroaniline was poorly visibleat the surface (Fig. 3), suggesting that it is better dis-

persed within the polymer matrix than piroxicam orrhodamine B.

Hydrolytic degradation of blanc microspheres

Blank microspheres obtained fromPSABESB90 and PSAGESB70 were subjected tohydrolytic degradation. The degradation of themicroparticles was run for 14 days in the PBS at37OC. The hydrolytic degradation was followedby determination of the weight loss of the sam-ples. All microspheres degraded completely toproducts soluble in water over two weeks where-as they lost nearly 90% of its initial weight with-in first 7 days of immersion in PBS. Microspheresobtained from PSAGESB90 degraded slower thanthose obtained from PSAGESB70. It was alsostated that small microspheres degraded slightlyfaster (especially in the first period) than the largeones (Fig. 4). The dependence of chemical copm-position of poly(ester-anhydride)s and size ofmicroparticles on rate of hydrolytic degradationcan influence release characteristics of suchmicrospheres. It is discussed in the next part ofthis article.

Table 2. Thermal properties of large microspheres (blank and loaded) made of poly(ester-anhydride)s.

Kind of Thermal properties

polymer model compound Tm [OC] ∆Hm [J/g]

- 79.8 85.0

p-Nitroaniline 74.2 77.9 PSAGESB90

Piroxicam 71.5 57.8

Rhodamine B 73.7 65.3

- 78.9 81.2

p-Nitroaniline 76.4 79.5 PSAGESB70

Piroxicam 71.0 55.3

Rhodamine B 72.9 61.8

Figure 3. Scanning electron microphotographs of large microspheres loaded with p-nitroaniline (A) made of PSAGESB90, (B) made ofPSAGESB70


Drug loading and in vitro drug release

Three model compounds exhibiting variousdegrees of water solubility (Fig. 1) were encapsulatedin poly(ester-anhydride) microspheres. The encapsu-lation efficiency was dependent on the kind of modelcompound as well as on the kind of copolymers (esterto anhydride ratio) and the size of microparticles(small or large ones) (Table 1). Piroxicam, having thelowest solubility in water showed encapsulation effi-ciency in the range 7ñ30%. The loading efficiency ofthis hydrophobic compound was higher in

PSAGESB70 than in PSAGESB90. The loading effi-ciency of piroxicam in microspheres obtained fromPSAGESB70 was independent on the size ofmicroparticles, whereas in microspheres obtainedfrom PSAGESB90 significant dependence on sizewas observed (small microparticles contained 2-foldless of piroxicam). The p-nitroaniline with mediumsolubility in water showed the highest encapsulationefficiency in the range 12ñ62%. It was found that p-nitroaniline loading as well as its encapsulation effi-ciency increased with increasing the size of the

Figure 4. Effect of polymer composition and size of microparticles on the hydrolytic degradation of blank microspheres in PBS at 37OC

Figure 5. Cumulative release profiles of rhodamine B from various microspheres as a function of time


microspheres. For PSAGESB70 containing more oli-goester fragments, the loading of p-nitroaniline wasonly slightly dependent on the size of microspheres,whereas for PSAGESB90 this dependency wasstrong (Table 1). The small microspheres obtainedfrom PSAGESB90 contained 5-fold less of p-nitroaniline than the large ones. These results indicat-ed better compatibility of piroxicam and p-nitroani-line with PSAGESB70 than with more hydrophobicand crystalline PSAGESB90. Compared to piroxi-cam and p-nitroaniline, which were better encapsu-lated in PSAGESB70, rhodamine B was more effi-ciently loaded in microspheres obtained fromPSAGESB90. Amount of rhodamine B encapsulat-

ed in PSAGESB70 microspheres was 2-3-fold small-er than in microspheres obtained from PSAGESB90.This hydrophilic substance was also better encapsu-lated in small particles than in larger ones, probablydue to their rapid precipitation. Encapsulation effi-ciency of rhodamine B was in the range 14-37%,however, the theoretical loading of rhodamine B wasonly 30 µg/mg (drug/polymer) compared to 100µg/mg for p-nitroaniline and piroxicam. Lower load-ing typically results in higher encapsulation efficien-cy, due to smaller concentration gradients driving thedrug out of the polymer/solvent droplets. Thus, therhodamine encapsulation efficiency cannot be direct-ly compared.

Figure 6. Cumulative release profiles of p-nitroaniline from various microspheres as a function of time

Figure 7. Cumulative release profiles of piroxicam from various microspheres as a function of time


The profiles of rhodamine B, p-nitroaniline andpiroxicam release from small or large microspherescomposed of two poly(ester-anhydride)s differedwith SBA and OSAGE content, are shown inFigures 5, 6 and 7, respectively. The effect of poly-mer composition and microsphere size on drugrelease was different for the tree model compounds.

Rhodamine B release from microspheres wasvery rapid and nearly complete within 24 h. The rateof release was faster than the rate of hydrolyticdegradation of blank microparticles. This

hydrophilic substance was released practically inde-pendently on microsphere size and polymer compo-sitions (Fig. 5), only for large microspheres obtainedfrom PSAGESB90 somewhat slower release wasobserved during the first 8 h.

Gradual release of p-nitroaniline was observedwithin 7 days (only small microspheres obtainedfrom PSAGESBA90 released p-nitroaniline slightlylonger) but nearly 60% of the initial amount of thiscompound was released within the first 2 days (Fig.6). The strong dependence of the rate of release on

Figure 8. The comparison the release profiles of p-nitroaniline with hydrolytic degradation of small (A) and large (B) microspheres


the size of microparticles as well as on polymercomposition was not observed also in this case. Thep-nitroaniline release profiles showed similar bursteffect (ca. 40% of cumulative release within the firstfew hours) for all studied microspheres followed bya time-lag when only a little amount of drug wasreleased. Longer time-lag (up to 4 days) wasobserved for microspheres obtained fromPSAGESB70 compared to 2 days lag period forPSAGESB90 microspheres. The p-nitroanilinerelease from small microparticles correlated withhydrolytic degradation of blank particles independ-ently on polymer composition (Fig. 8A), but therelease profile from the large particles correlatedwith hydrolytic degradation profile only forPSAGESB70 (Fig. 8B). The release of p-nitroani-line from large microspheres obtained fromPSAGESB90 was faster than their hydrolytic degra-dation (Fig. 8B), probably due to better compatibil-

ity of p-nitroaniline with PSAGESB70 than withPSAGESB90.

Piroxicam release was the slowest one and incontrast to p-nitroaniline or rhodamine B, variedsignificantly with polymer composition and micros-phere size. Generally, it could be stated that thesmall microspheres released the piroxicam slowerthan the large ones (bellow 80% of piroxicam wasreleased within 14 days from small particles, where-as above 90% from larger ones). During the first 4days, in which 53-65% of this model compound wasreleased, the rate of release was higher for micro-spheres composed from PSAGESB70 (Fig. 7). Thepiroxicam release profiles showed a burst effect fol-lowed by a short Ñlagî period and next more rapidrelease phase between 2-7 days. Small microparti-cles exhibited much smaller burst effect than thelarge ones. Below 10% of cumulative release wasobserved within the first 8 h for small PSAGESB90

Table 3. Kinetic parameters (n and K) estimated for Korsmeyer-Peppas model for release of different model compounds from variousmicrospheres during the first period (ca. 60% release) and for total release.

Kind of 60 % of releaseA) total releaseB)

model compound

microspheres n K r2 C) n k r2 C)

PSAGESB90 large

0.70 21.18 0.98 0.49 24.83 0.88

PSAGESB70 large

0.44 43.10 0.94 0.27 48.11 0.81Rhodamine B

PSAGESB90small

0.42 43.98 0.85 0.27 49.33 0.73

PSAGESB70small

0.31 50.91 0.98 0.21 54.64 0.86

PSAGESB90large

0.23 58.26 0.89 0.23 58.30 0.94

PSAGESB70large

0.25 57.21 0.95 0.23 55.5 0.97p-Nitroaniline

PSAGESB90small

0.31 53.12 0.91 0.28 49.91 0.95

PSAGESB70small

0.36 54.71 0.95 0.31 49.55 0.95

PSAGESB90large

0.28 37.56 0.93 0.30 39.29 0.93

PSAGESB70large

0.22 45.98 0.91 0.23 46.85 0.96Piroxicam

PSAGESB90small

0.67 14.31 0.94 0.66 14.40 0.96

PSAGESB70small

0.46 30.24 0.93 0.43 29.90 0.97

A) 60% of release (4 days for piroxicam, 2 days for p-nitroaniline and 8 h for rhodamine B); B) total release (30 days for piroxicam, 14 daysfor p-nitroaniline and 24 h for rhodamine B); C) r2 ñ correlation coefficient.


microspheres, whereas nearly 40% of piroxicamwas released during that period from largePSAGESB70 microspheres. The release of piroxi-cam from large microparticles was better correlatedwith hydrolytic degradation of blank particles thanthe release of this compound from small micro-spheres (Fig. 9). The release of piroxicam fromsmall particles was slower than hydrolytic degrada-tion of blank microspheres (Fig. 9A). It is probablydue to more hydrophobic nature of piroxicam-loaded particles compared to blank ones.

Kinetics of model compound release

The mechanism of drug release may be studiedby using different release models. Modeling of the

controlled release of drugs from polymeric deviceshas been a subject of considerable research over thepast 30 years. Higuchi proposed (24) a simple rela-tionship that described drug release from a matrix asa function of time (eq. 5).

Mt/M8 = K t1/2 (5)where Mt/M8 is the amount of released drug (%) attime t. The relative release of drug is proportional tothe square root of time. Korsmeyer and Peppas (21,25) extended the Higuchi model to a more generalform (eq. 6).

Mt/M8 = K tn (6)where K is a kinetic constant incorporating structuraland geometrical characteristics of the drug-loadeddevices, and n is the release exponent, indicative of

Figure 9. The comparison the release profiles of piroxicam with hydrolytic degradation of small (A) and large (B) microspheres


the mechanism of drug release. This model is used toanalyze the release from polymeric dosage formswhen the release mechanism is not well known orwhen more than one type of release phenomena areinvolved. For drug-loaded slab or film system, valuesof n = 0.5 or less correspond to a purely Fickian dif-fusion mechanism (18). When n = 1.0, the rate ofrelease is independent on time and the drug releasefollows zero-order kinetics. The value of n between0.5 to 1.0 is an indicator of the superposition of twoindependent mechanism of drug transport (anom-alous transport) (26, 27). For a spherical drug carriers,the threshold of the n-value distinguishing betweenFickian and non-Fickian diffusion mechanism hasbeen slightly modified and thus n-values between0.43 and 0.85 represent anomalous transport (28).

Kinetic parameters K and n for microspheresloaded with three different model compounds werecalculated and summarized in Table 3. The kineticparameters were calculated for total release profileand for first period in which ca. 60% of model com-pound was released. The equation (6) is valid for thefirst 60% of the release, especially when the diffu-sion play the important role in release mechanism(29). As seen in Table 3, the n-values were higherthan 0.43 for large microspheres loaded with rho-damine B and for small microspheres loaded withpiroxicam. It suggested that the release of a low-molecular weight hydrophilic compounds fromlarge microparticles and low-molecular weighthydrophobic compounds from small ones followedanomalous transport mechanism related to non-Fickian model. The n-value was higher for micros-pheres obtained from PSAGESB90 containing moreanhydride than ester fragments. It indicated releasemechanism closer to zero order release kinetics. Theresults achieved in this study indicated that the sizeof microspheres as well as polymer composition canplay the important role in release mechanism ofhydrophilic or hydrophobic drugs from poly(ester-anhydride) microspheres.

It was stated that for microspheres containingp-nitroaniline, a low-molecular weight compoundwith medium hydrophilicity, the n-values werebelow 0.43, suggesting that the release mechanismof such compounds is Fickian-diffusion controlled.The kinetic parameters (n and K-values) estimatedfor Korsmeyer-Peppas model for all microparticlesloaded with p-nitroaniline were similar, independ-ently on size of microparticles and composition ofthe polymer. It correlated with release profiles of p-nitroaniline showed in Figure 6.

Those results indicated that the release ofencapsulated drugs from poly(ester-anhydride)

microspheres depended on kind of encapsulateddrugs, chemical composition of polymers (due todifferent compatibility of drug with polymer matrixand due to various rate of hydrolysis of polymer),and also on particle size. Therefore, by adjustingthese parameters it is possible to obtain differentrelease rates and profiles, which might be of interestfor different therapeutic purposes and modalities ofadministration.

CONCLUSIONS

The present study demonstrates that the emul-sion solvent evaporation method can be useful toformulate small or large microspheres from func-tional poly(ester-anhydride)s, loaded with modelcompounds with different hydrophilicity. The kindof model compound loaded influenced size and sur-face morphology of microparticles obtained. Themain objective of this study was to evaluate theinfluence of polymer composition (OSAGE andSBA content) and size of microspheres on loadingefficiency and release profiles of model compounds.The highest encapsulation efficiency was obtainedfor p-nitroaniline, probably due to its good compat-ibility with polymers. Piroxicam and p-nitroaniline,(hydrophobic compounds) were better encapsulatedin PSAGESB70,whereas rhodamine B (hydrophiliccompound) was more efficiently loaded inPSAGESB90 microspheres. The piroxicam loadedmicrospheres exhibit the most interesting releaseprofiles, showing that elongation of release as wellas reduction of burst effect can be achieved bydecreasing microsphere size. The release ofhydrophobic compounds from small poly(ester-anhydride) microparticles followed anomaloustransport mechanism, for PSAGESB70, closer toFickian diffusion and for PSAGESB90 closer tozero order release kinetics. When rhodamine B or p-nitroaniline were encapsulated, it may be difficult totailor release profiles by controlling microspheresize and polymer composition. An understanding ofthe factors affecting drug release mechanisms frompoly(ester-anhydride) microspheres allow to predictthe behavior of biologically active substances, even-tually incorporated into microparticulate systemsand to design particular polymer-drug systems forvarious purposes.

Acknowledgment

The author thanks the Polish State Committeefor Scientific Research for financial support by grantno. N205 016534.


Declaration of interest

The author reports no conflicts of interest. Theauthor alone is responsible for the content and writ-ing of this article.

REFERENCES

1. Freiberg S., Zhu, X.: Int. J. Pharm. 282, 1(2004).

2. Hoffman A.S.: J. Control. Release 132, 153(2008).

3. Kim S., Kim J.H., Jeon O., Kwon I.C., Park K.:Eur. J. Pharm. Biopharm. 71, 20 (2009).

4. Slomkowski S.: Macromol. Symp. 288, 121(2010).

5. Kipper M.J., Shen E.B.: Biomaterials 23, 4405(2002).

6. Tran V.T., Benoit J.P., Venier-Julienne M.C.:Int. J. Pharm. 407, 1 (2011).

7. Berkland C., Kim K., Pack D.W.: Pharm. Res.20, 1055 (2003).

8. Berkland C., Kipper M.J., Narasimhan B.,Kyekyoon K.K., Pack D.W.: J. Control. Release94, 129 (2004).

9. Dawes G.J., Fratila-Apachitei L.E., Mulia K.,Apachitei I., Witkamp G.J., Duszczyk J.: J.Mater. Sci. Mater. Med. 20, 1089 (2009).

10. Bezemer J.M., Radersma R., Grijpma D.W.,Dijkstra P.J., Blitterswijk C.A., Feijen J.: J.Control. Release 67, 249 (2000).

11. Modi S., Jain J.P., Domb A.J., Kumar N.: Eur.J. Pharm. Biopharm. 64, 277 (2006).

12. Tamada J., Langer R.: Proc. Natl. Acad. Sci.USA 90, 552 (1993).

13. Gopferich A.: Biomaterials 17, 103 (1996).

14. Wang X.P., Chen T.N., Yang Z.X.: Sens.Actuators A 133, 363 (2007).

15. Pfeifer B.A., Burdick J.A., Langer R.:Biomaterials 26, 117 (2005).

16. Shen E., Kipper M.J., Dziadul B., Lim N.K.,Narasimhan B.: J. Control. Release 82, 115(2002).

17. Ferreira P., Almeida A.J.: J. Control. Release97, 431 (2004).

18. Bettencourt A., Florindo H.F., Ferreira I.F.S.,Matos A., Monteiro J., Neves C., Lopes P.,Calado A., Castro M., Almeida A.J.: J.Microencapsul. 27, 533 (2010).

19. Jaszcz K.: Int. J. Polym. Mater. Polym. Bio-mater. in press (2013).

20. Jaszcz K.: Macromol. Symp. 254, 109 (2007).21. Korsmeyer R.W., Gurny R., Doelker E.M., Buri

P., Peppas N.A.: Int. J. Pharm. 15, 25 (1983).22. Peppas N.A.: Pharm. Acta Helv. 60, 110

(1985).23. Jaszcz K., £ukaszczyk J.: Polym. Degrad.

Stabil. 96, 1973 (2011).24. Higuchi T.: J. Pharm. Sci. 50, 874 (1961).25. Ritger P.L., Peppas N.A.: J. Control. Release 5,

23 (1987).26. Kumar N., Muzzarelli R., Muzzarelli C.,

Sashiwa H., Domb A.J.: Chem. Rev. 104, 6017(2004).

27. Shelke N.B., Aminabhavi T.M.: Int. J. Pharm.,345, 51 (2007).

28. Wu J., Liao C., Wang Z., Cheng W., Zhou N.,Wang S., Wan Y.: React. Funct. Polym. 71, 925(2011).

29. Vasheghani-Farahani E., Khorram M.: J. Appl.Polym. Sci. 83, 1457 (2002).

Received: 22. 01. 2013

in vitrorelease of model compounds with different ... · the procedure for preparation of...

Documents